Wire size?

alyaz · February 2013

Can any of you geniuses (meant in the most complimentary way :-) give me your take on what wire size and size breaker should be used for the following application? Please? I live in an area where I need to have my facts in front of me when getting suggestions/quotes from the one or maybe two installers available...

Six Helios 250 Watt panels;
Two parallel runs of three panels in series;
80 feet between panels to Midnite Classic 150 Controller.

Helios Panel Specs:
MPP Voltage (V) 30.3
MPP Current (A) 8.22
Open Circuit (V) 37.4
Short Circuit (A) 8.72

Anything more needed? :-)

westbranch · February 2013

Re: Wire size?

A combiner box, weather proof before the CC, can be at the array or ?? Decide on location based on line loss (depends on wire size used), & if you need 1 or 2 sets from the combiner to CC.

alyaz · February 2013

Re: Wire size?

I am hoping to use a weather proof combiner box near the array. The combiner box will be about 70 feet from the Midnite Classic 150 Controller. When I used the Midnite Classic String Sizing Tool on their site, it seems the best option is to run two sets of three panels in series. Our battery bank is a 24 V system / 1065 a/hrs.

NorthGuy · February 2013

Re: Wire size?

With 2 series, you'll have 17A of current in your wires. The minimum wire size for this is #10.

With 160 feet of #10 wire, you will lose 46W when your panels are producing at their max, or 11W when they're producing 50% of nominal.

If you want to reduce these losses, you can use thicker wire.

alyaz · February 2013

Re: Wire size?

Sorry, can you help me understand the wiring itself... do the two series strings not combine with mc4's to the junction box (which is at the array). So wouldn't you just have one 70 foot run of wire from the array combiner to the controller?

niel · February 2013

Re: Wire size?

you are correct as you will get 90.9v at 8.22a per string of pvs. paralleling 2 strings will still have the 90.9v, but the currents will add giving you 16.44a to go down a single pair of wires from the combiner. higher currents equals higher losses and so the higher current does mean the wire size must be increased to reduce v drop losses. no matter if you run a single pair (1+ and 1-) from the combiner to the cc or run 2 pair (2+ and 2-) back to the cc to be combined there, you will still need the same amount of copper presenting the same amount of resistance in total in both cases.

as an example if each string of pvs were to be run back to the cc to be combined there and each were #10 wire then to combine at the pvs first you will need to double the area of copper and that is equal to moving 3 places on the awg to #7 to present the same resistance and v drop loss as the individual strings using #10. #7 is not a common number so one would most likely go to #6 to make it even better.

YehoshuaAgapao · February 2013

Re: Wire size?

NorthGuy wrote: »

With 2 series, you'll have 17A of current in your wires. The minimum wire size for this is #10.

With 160 feet of #10 wire, you will lose 46W when your panels are producing at their max, or 11W when they're producing 50% of nominal.

If you want to reduce these losses, you can use thicker wire.

For finding the correct wiring and conduit size, and especially over-engineering it, this is your best friend: http://www.electriciancalculators.com/ - download the code calculators VB app

Most people will use #10AWG pre-combiner and #6AWG-#2AWG post-combiner (depending on wire distance). The state of california sample design had the combiner very close to the array but ran 1/0 from the combiner to the inverter.

If you want grid-tie level wiring loss efficiency in off-grid you will need #6AWG pre-combiner and 1/0AWG post-combiner (assuming 4 strings), which is the maximum wire size for both input and output for the Midnite Solar MPNV6 combiner box.

I'm using #6/#2 because I put the combiners closer to the inverter - large conduit is expensive and difficult to deal with. Up to 90 ft from panel to inverter in #6, 25ft from combiner to PDP in #2. #6 can be run in 3/4" which is still cheap (bend it yourself). Like California sample design but most of the distance is pre-combiner instead of post-combiner (James likes the combiners in the shade, plus his arrangement is expansion-friendly). 1" and over things start getitng expensive (prefab). Over 2" things get REALLY expensive (prefab outrageously expensive if even available, 2.5" usually is). #2 AWG is the temperature-compensated code minimum for 4 strings in 65C, 118 degree summer days in Phoenix plus 15C for being in sun - 50C ambient adding 15C for sun-struck conduit, 4 combiner box output circuits combined into a single conduit to use the rear 2.5" knockout in the PDP. Most electricians/engineers do not temperature-derate. Xantrex does not either - 2 circuits in conduit, 60A continuous, hotter than 35C, #6AWG is insufficient. Was forced to use #6 anyway because there aren't enough knockouts in the PDP to do 1 circuit per conduit and James was very disenchanted in the tight wiring compartments in the charge controllers with the #6, he refused to use anything bigger - he is very anal about wire management - bend radiuses, rubbing on stuff, etc... . The grounding wire management is even worse in those things...

Oh as a rule of thumb, moving up 3 wire sizes doubles the wire bandwidth - #7 is double of #10 (round to #6), #4 quadruple of #10; #3 double of #6 (round to #2), 1/0 quadruple of #6.

alyaz · February 2013

Re: Wire size?

Thanks guys. And thanks niel, that description helps even me understand it! :-)

NorthGuy · February 2013

Re: Wire size?

YehoshuaAgapao wrote: »

Most people will use #10AWG pre-combiner and #6AWG-#2AWG post-combiner (depending on wire distance). The state of california sample design had the combiner very close to the array but ran 1/0 from the combiner to the inverter.

It depends on the size and configuaration of solar array, distance from combiner box to the controller, and other factors. Based on these factors, you choose the wire size that meets your needs, and it will be different for different situations. If someone used #2 wire in some situation, doesn't mean it should be used everywhere.

If cost is not a concern, thicker wire is always better.

Smaller wire is a problem because it dissipates more heat. There are two bad consequences of that heat.

First, wire can overheat, which can damage insulation or even destroy the wire. This depends on the current going through the wire, insulation, wire packing in the conduit, ambient temperature etc. Instead of considering all these factors, you can look-up NEC tables which give you the minimum safe wire size for given current. In this case, 17A current require at least #10 wire. These tables are very conservative. If they say #10 can carry up to 20A, you can consider #10 to be safe for the task.

Second, heat losses, which depend on wire size and lenfth. You can get thicker (more expensive) wire and get less losses, or you can use smaller (less expensive) wire and get more losses. The loss can be calculated as I^2*R, where I is the current, and R is the wire resistance. Wire resistance depends on the wire size and its lenfth. Looking at the AWG tables for #10 wire, you can find out that its resistance is about 1mOhm/foot. If you need 160ft (80ft forth and 80ft back), the resistance will be 160mOhm. At 17A total losses will be 17*17*0.160 = 46W. For #6 wire, which has resistance of 0.4mOhm/ft, the loss would be 18W. With #2 wire the loss is 7W. It is up to the designer to select an optimum solution based on projected losses and the cost of wire.

vtmaps · February 2013

Re: Wire size?

Your choices are
three parallel strings of two panels per string (Vmp = 60.6)
or
two parallel strings of three panels per string (Vmp = 90.9)

As with most things the choice is not simple. Your charge controller will work more efficiently, and therefore run cooler and last longer if you choose the lower voltage (60.6). We don't have efficiency curves for the Classic charge controller, but we do have them for the Outback controller. The Classic was designed by the same engineers who designed the Outback controller and they have indicated (on this and other forums) that the efficiency of the Classic (like the Outback) is higher at the lower input voltage.

If you choose to configure your panels at the higher voltage, the advantage is that you can get away with spending a bit less on the 80 ft cable.

Lets take an example:

If you do two strings (Vmp = 90.9) and run the combined output 80 ft through #8 gauge wire, you will have a 1.85% voltage drop in the cable, representing a 28 watt power loss at full nameplate power (seldom achieved).

If you do three strings (Vmp = 60.6) and run the combined output 80 ft through #6 gauge wire, you will have a 2.62% voltage drop in the cable, representing a 39 watt power loss at full nameplate power (seldom achieved).

But let's assume that the controller efficiency is 1% better at the lower input voltage. At full power of 1500 watts (seldom achieved) the higher input voltage (90.9) will cause the controller to burn an additional 15 watts.

Therefore, at the higher voltage you will lose 28 watts in the cable and an extra 15 watts in the controller. At the lower voltage you will lose 39 watts in the cable and nothing extra in the controller.

Furthermore, The cable losses go as the square of the current. At half power the cable losses are one quarter of the above numbers. The difference in efficiency of the controller (at the two input voltages) may actually increase a bit at half power. Therefore, when you are operating at less than full power (which is typical) the advantage shifts towards the lower input voltage.

The decision often comes down to whether you would rather create the extra heat in the cable or in the controller. To see another example of this analysis look here: http://forum.solar-electric.com/showthread.php?15907

Bottom line: If you can afford 80 ft of #6 cable, my advice is to configure three parallel strings with two panels per string (Vmp = 60.6).

--vtMaps

NorthGuy · February 2013

Re: Wire size?

I have one more consideration.

You may to decide to expand your system in the future. If there's a room for more panels, it's good idea to size the wire for a bigger system (which means using thicker wire). This way, you do not need to re-wire when you add more panels.

alyaz · February 2013

Re: Wire size?

vtmaps wrote: »

Your choices are
three parallel strings of two panels per string (Vmp = 60.6)
or
two parallel strings of three panels per string (Vmp = 90.9)

As with most things the choice is not simple. Your charge controller will work more efficiently, and therefore run cooler and last longer if you choose the lower voltage (60.6). We don't have efficiency curves for the Classic charge controller, but we do have them for the Outback controller. The Classic was designed by the same engineers who designed the Outback controller and they have indicated (on this and other forums) that the efficiency of the Classic (like the Outback) is higher at the lower input voltage.

If you choose to configure your panels at the higher voltage, the advantage is that you can get away with spending a bit less on the 80 ft cable.

Lets take an example:

If you do two strings (Vmp = 90.9) and run the combined output 80 ft through #8 gauge wire, you will have a 1.85% voltage drop in the cable, representing a 28 watt power loss at full nameplate power (seldom achieved).

If you do three strings (Vmp = 60.6) and run the combined output 80 ft through #6 gauge wire, you will have a 2.62% voltage drop in the cable, representing a 39 watt power loss at full nameplate power (seldom achieved).

But let's assume that the controller efficiency is 1% better at the lower input voltage. At full power of 1500 watts (seldom achieved) the higher input voltage (90.9) will cause the controller to burn an additional 15 watts.

Therefore, at the higher voltage you will lose 28 watts in the cable and an extra 15 watts in the controller. At the lower voltage you will lose 39 watts in the cable and nothing extra in the controller.

Furthermore, The cable losses go as the square of the current. At half power the cable losses are one quarter of the above numbers. The difference in efficiency of the controller (at the two input voltages) may actually increase a bit at half power. Therefore, when you are operating at less than full power (which is typical) the advantage shifts towards the lower input voltage.

The decision often comes down to whether you would rather create the extra heat in the cable or in the controller. To see another example of this analysis look here: http://forum.solar-electric.com/showthread.php?15907

Bottom line: If you can afford 80 ft of #6 cable, my advice is to configure three parallel strings with two panels per string (Vmp = 60.6).

--vtMaps

Wow, thanks for that. You'd be a killer salesperson... Got me pursuaded. Thanks.

alyaz · February 2013

Re: Wire size?

Just thinking... From changing my plan to go from two strings to three strings, I will now need fuses on all three strings, correct? So that would require a disconnect box/fuses, right?

BB. · February 2013

Re: Wire size?

That is usually correct... You need to look for the series fuse rating on the solar panel you plan on using to be sure.

-Bill

vtmaps · February 2013

Re: Wire size?

alyaz wrote: »

Just thinking... From changing my plan to go from two strings to three strings, I will now need fuses on all three strings, correct? So that would require a disconnect box/fuses, right?

The box you want is a combiner box. I recommend circuit breakers rather than fuses. Midnite makes a nice three circuit combiner box. They also make combiner boxes with built-in MC4 connectors and fuses. Its not in the catalog, but they will custom make one with circuit breakers and built-in MC4 connectors.

--vtMaps

alyaz · February 2013

Re: Wire size?

Copy, thanks very much.

YehoshuaAgapao · February 2013

Re: Wire size?

It don't hurt to go to MPNV6, for expansion-friendliness, if you add more strings later that have similar tilt/orientation/shading/voltage (60-cell, 72-cell, 80-cell, 96-cell) properties.

vtmaps · February 2013

Re: Wire size?

YehoshuaAgapao wrote: »

It don't hurt to go to MPNV6, for expansion-friendliness

Not a bad idea, but if expansion is in the future this is more important:

NorthGuy wrote:

If there's a room for more panels, it's good idea to size the wire for a bigger system (which means using thicker wire).

Its a lot easier to upgrade your combiner in the future than to upgrade your 80 ft cable. A larger combiner may be of value in the future, but a larger cable will be of value now and in the future.

--vtMaps

YehoshuaAgapao · February 2013

Re: Wire size?

Its a lot easier to upgrade your combiner in the future than to upgrade your 80 ft cable. A larger combiner may be of value in the future, but a larger cable will be of value now and in the future.

Very true. I did #2 on all my combiners because they will all eventually have 4 strings on them. Currently I have 1,2,2,4 strings (4 combiners). At expansion, controllers #1 and 2 will be merged and get 1 string added (the standoffs successfully raised the panels above the shade cast by the air conditioner - no shading even at winter solstice). Controller #3 will get 2 strings added. Controller #2 (freed up from the merge) will get a 12-panel ground or pole (2 wind turbine masts, depending on $$$) mount. The labor savings will be substantial as bigger wire will need bigger conduit, plus long runs are labor intensive when fishing. It don't hurt to oversize conduit by one size level either - makes the wire pull a lot easier especially in flex or when there are more than two bends.

Wire size?

Comments

Categories